Evidence of growth-selective predation on larval Japanese anchovy Engraulis japonicus in Sagami Bay

Growth-selective predation mortality was demonstrated for postlarval Japanese anchovy Engraulis japonicus in field research. The larval anchovy and their predatory fish were simultaneously captured by a trawl in Sagami Bay during October to November 2000. The growth rates analyzed by otolith microstructure were compared between the larvae from the stomach contents of the predators (prey larvae) and those from the population of origin (surviving larvae). The mean growth rates of the prey larvae collected on 28 October and 2 to 4 November (mean +/- SD: 0.57 +/- 0.07 mm d(-1)) and on 23 November (0.50 +/- 0.06 mm d(-1)) were significantly lower than those of the corresponding surviving larvae (0.63 +/- 0.07 and 0.54 +/- 0.06 min d(-1), respectively). Such significant differences were not explained by size-selective predation, but were due to variations in the mean growth rates at the same larval size (i.e. non-size-related). The mean growth rates of the prey larvae were different among predatory species (barracuda Sphyraena pinguis, Japanese sea bass Lateolabrax japonicus, white croaker Pennahia argentatus, Japanese jack mackerel Trachurus japonicus, Pacific round herring Etrumeus teres and juvenile anchovy). Comparisons of back-calculated daily growth rates showed that the decrease in growth rates of the prey larvae were consistent from directly after hatching up to predation. The larvae with lower growth rates were more vulnerable to predation, owing to the cumulative decline in growth rates from hatching to each encounter with predators, compared to the larvae with higher growth rates, even if they were the same size, at a given moment in the sea. Therefore, the level of growth rates itself had direct impact on vulnerability to predation for larval anchovy, independently of both size (size-selective mortality) and time (stage duration). In addition, such impacts could be predator specific. We propose the 'growth-selective predation' hypothesis (mechanism), which is theoretically independent of and synergistic with the existing hypotheses based on size and time under the general 'growth-mortality' concept for the survival process during the early life history of marine pelagic fish.